Whisper
Open-source speech recognition model from OpenAI that transcribes and translates audio in dozens of languages.
About
Whisper is a general-purpose speech recognition model released as open source by OpenAI. It was trained on a large and diverse dataset of audio, giving it strong accuracy across many languages, accents, and recording conditions. It handles transcription, language detection, and speech translation in a single model.
Whisper comes in six sizes ranging from tiny to large, letting developers choose the right balance of speed and accuracy for their use case. It runs via a command-line interface or Python API and integrates into any project with standard tools. Because it is released under the MIT License, there are no licensing fees; you only pay for the compute you use when self-hosting.
Whisper comes in six sizes ranging from tiny to large, letting developers choose the right balance of speed and accuracy for their use case. It runs via a command-line interface or Python API and integrates into any project with standard tools. Because it is released under the MIT License, there are no licensing fees; you only pay for the compute you use when self-hosting.